Pelvic and lumbar splanchnic afferents may play an important role in mediating both noxious and non-noxious sensation from the colon. These afferents have been categorized previously by electrophysiological studies as three populations, according to the location of their receptive fields: mucosal afferents, muscular afferents and serosal afferents (Lynn & Blackshaw, 1999). No information is available on the structure of any of these endings. We aimed to correlate the function and structure of serosal afferent endings using a combination of electrophysiological recording and a modified anterograde labelling technique. The distal colon with attached inferior mesenteric ganglion (IMG) and lumbar splanchnic nerves (LSN) were removed from fifteen humanely killed rats and perfused in an organ bath with modified Krebs solution. The LSN was drawn into a second paraffin-filled chamber and small strands of fibres placed onto an electrode for recording. Receptive fields were located by mechanical stimulation. The recorded strand was treated with a 1Ð5 % biotinamide tracer in artificial intracellular solution (Tassicker et al. 1999) for 20 h at 35Ð37 °C. The tissue was then removed and fixed, then treated with extravidin-FITC for visualization with a fluorescence microscope. Histochemistry was also performed to visualize isolectin B4 (IB4) binding and calcitonin gene-related peptide (CGRP) immunoreactivity (specific for sensory fibres). Serosal afferents were characterized by their potent response to blunt probing with a glass rod, and their lack of response to stretch of the colonic muscle, or fine stroking of the mucosa. Six strands that contained active serosal afferents were treated with biotinamide tracer after accurate mapping of their receptive field. Endings of four afferents were observable with anterograde tracing, and two were not, but these could be visualised with IB4 binding. Anatomical assessment showed endings in the mesentery within 100 µm of vascular smooth muscle. They appeared consistently as spray-like endings occupying an area < 300 µm in diameter. They were not associated with any specialized non-neuronal tissue. Although IB4 consistently labelled mesenteric fibres, it did not label endings in the muscular layers. CGRP immunohistochemistry, however, revealed intraganglionic laminar endings in the myenteric plexus and intramuscular arrays in circular muscle that are known to belong to mechanoreceptors (Zagorodnyuk & Brookes, 2000). We presume these to correspond to stretch-sensitive muscular afferents. CGRP labelling in mucosa was sparse but clearly observed surrounding crypts. IB4 labelling in mucosa was inconclusive due to endothelial staining. In conclusion, we have generated a working model for the correlation of structure and function in mechanoreceptors of the colon, and provided evidence for specific chemical coding of two subpopulations.

This work was funded by NHMRC project grant no. 104814.

Experiments were carried out with the approval and under the guidelines of the Ethics Committee of the IMVS, Adelaide.